U.S. patent application number 13/318162 was filed with the patent office on 2012-02-23 for external defibrillator.
Invention is credited to John McCune Anderson, Rebecca Dimaio, Cesar Oswaldo Navarro-Paredes.
Application Number | 20120046706 13/318162 |
Document ID | / |
Family ID | 42288516 |
Filed Date | 2012-02-23 |
United States Patent
Application |
20120046706 |
Kind Code |
A1 |
Anderson; John McCune ; et
al. |
February 23, 2012 |
EXTERNAL DEFIBRILLATOR
Abstract
An external defibrillator estimates the phase of ventricular
defibrillation (VF) by deriving, from an ECG exhibiting VF, at
least one quality marker representing the morphology of the ECG
and, therefore, the duration of the VF. The duration of the VF is
calculated as a function of the value (s) of the quality marker
(s). The quality marker (s) may comprise any one or more of the
median slope of the ECG, the average slope of the ECG, the ratio of
the power in relatively high and low frequency bands of the ECG,
and a measure of the density and amplitude of peaks in the ECG,
over a predetermined period.
Inventors: |
Anderson; John McCune;
(Holywood, GB) ; Navarro-Paredes; Cesar Oswaldo;
(Whiteabbey, GB) ; Dimaio; Rebecca; (Belfast,
GB) |
Family ID: |
42288516 |
Appl. No.: |
13/318162 |
Filed: |
April 28, 2010 |
PCT Filed: |
April 28, 2010 |
PCT NO: |
PCT/EP2010/055742 |
371 Date: |
October 31, 2011 |
Current U.S.
Class: |
607/5 |
Current CPC
Class: |
G06K 9/0053 20130101;
A61B 5/361 20210101; A61B 5/316 20210101; A61N 1/3925 20130101 |
Class at
Publication: |
607/5 |
International
Class: |
A61N 1/39 20060101
A61N001/39 |
Foreign Application Data
Date |
Code |
Application Number |
May 1, 2009 |
IE |
S2009/0348 |
Claims
1. An external defibrillator comprising means for estimating the
phase of ventricular fibrillation (VF) by analysis of the patient's
ECG and means dependent on the estimated phase for indicating
whether an immediate shock or CPR is advised.
2. A defibrillator as claimed in claim 1, wherein the means for
estimating the phase of VF comprises estimating VF duration and
comparing the estimated duration with a threshold level.
3. A defibrillator as claimed in claim 2, wherein VF duration is
estimated by deriving at least one VF quality marker from a
patient's ECG and calculating the duration of VF as a function of
the value(s) of the quality marker(s).
4. A defibrillator as claimed in claim 3, wherein the quality
marker comprises the median slope of the ECG over a predetermined
period.
5. A defibrillator as claimed in claim 3, wherein the quality
marker comprises the average slope of the ECG over a predetermined
period.
6. A defibrillator as claimed in claim 3, wherein the quality
marker comprises the ratio of the power in relatively high and low
frequency bands of the ECG over a predetermined period.
7. A defibrillator as claimed in claim 1, wherein the quality
marker comprises a measure of the density and amplitude of peaks in
the ECG over a predetermined period.
8. A defibrillator as claimed in claim 1, wherein the duration of
VF as a function of the value(s) of the quality marker(s) is
calculated as: t = - B + C [ - ln ( Q - D A ) ] E ##EQU00010##
where t is the estimated VF duration in seconds, A, B, C, D and E
are constants, and Q is the value of the quality marker, if only
one, or a linear combination the values of the quality markers if
more than one.
9. A defibrillator as claimed in claim 1, wherein the means for
estimating the phase of VF comprises deriving a quantity related to
the density and amplitude of peaks in the ECG over a predetermined
period.
10. A defibrillator as claimed in claim 9, wherein said quantity is
derived by constructing an envelope of the ECG and measuring the
average magnitude of peaks lying above the envelope during the
predetermined period.
11. A defibrillator as claimed in claim 10, wherein the said
quantity is compared to a threshold level to estimate the phase of
VF.
Description
FIELD OF THE INVENTION
[0001] This invention relates to an external defibrillator.
BACKGROUND
[0002] Following the publication of a three-phase time sensitive
model by Weisfeldt and Becker (Weisfeldt M L, Becker L B:
"Resuscitation after cardiac arrest. A 3-phase time-sensitive
model"; JAMA. 2002; 288: 3035-3038), much research has focused on
developing treatment algorithms specific to each of three phases of
cardiac arrest.
[0003] The first phase is known as the "Electrical Phase" and
constitutes the first four minutes of a cardiac arrest. During this
time, immediate defibrillation should be administered.
[0004] The second phase is known as the "Circulatory Phase" and
occurs after the first phase for another period of four
minutes--that is, four to ten minutes after arrest. During this
phase, CPR should be administered before defibrillation in order to
increase perfusion and prepare the myocardium for defibrillation by
re-oxygenation, thereby increasing the chances of success of the
therapy.
[0005] The final phase is known as the "Metabolic Phase" and the
only available treatments are mild or moderate hypothermia,
metabolic therapies or the use of Caspase inhibitors, all of which
are only applicable to in-hospital patients.
[0006] Studies have shown that survival rates are much lower for
patients presenting prolonged ventricular fibrillation (VF). In
these cases, immediate defibrillation appears to simply convert the
patient's electrocardiogram (ECG) from one non-perfusing rhythm
(i.e. VF) to another (i.e. PEA/asystole). It has also been shown
that immediate defibrillation in cases of prolonged cardiac arrest
would result from countershock-induced injury to ischemic
myocardium.
[0007] The condition of the myocardium deteriorates rapidly without
effective CPR to perfuse the heart muscle and other vital organs.
It is widely accepted that, for VF of short duration (less than 4
mins since VF onset), immediate shock therapy is indicated, whereas
for VF of long duration (more than 4 mins since VF onset), CPR
prior to defibrillation increases the chances of return of
spontaneous circulation (ROSC).
[0008] It is evident that if the responder had accurate information
as to which phase of VF the patient was presenting, they could
deliver the most appropriate form of therapy and improve their
chances of survival.
SUMMARY OF THE INVENTION
[0009] According to the present invention there is provided an
external defibrillator comprising means for estimating the phase of
ventricular fibrillation (VF) by analysis of the patient's ECG and
means dependent on the estimated phase for indicating whether an
immediate shock or CPR is advised.
[0010] In certain embodiments the means for estimating the phase of
VF comprises estimating VF duration and comparing the estimated
duration with a threshold level. In such a case VF duration is
preferably estimated by deriving at least one VF quality marker
from a patient's ECG and calculating the duration of VF as a
function of the value(s) of the quality marker(s).
[0011] The quality marker may comprise the median slope of the ECG
over a predetermined period, the average slope of the ECG over a
predetermined period, the ratio of the power in relatively high and
low frequency bands of the ECG over a predetermined period, or a
measure of the density and amplitude of peaks in the ECG over a
predetermined period.
[0012] In another embodiment the means for estimating the phase of
VF comprises deriving a quantity related to the density and
amplitude of peaks in the ECG over a predetermined period.
[0013] In such case said quantity is derived by constructing an
envelope of the ECG and measuring the average magnitude of peaks
lying above the envelope during the predetermined period. The said
quantity is preferably compared to a threshold level to estimate
the phase of VF.
BRIEF DESCRIPTION OF THE DRAWINGS
[0014] FIG. 1 shows how VF changes its morphology from its onset
with time.
[0015] FIG. 2 shows the construction of an ECG envelope in one
embodiment of the invention.
[0016] FIG. 3 is the flow diagram of an algorithm to calculate the
envelope of FIG. 2 for use in an embodiment of the invention.
[0017] FIG. 4 is a block diagram of an automated external
defibrillator embodying the invention.
[0018] FIG. 5 is a flow diagram of a further embodiment.
DESCRIPTION OF EMBODIMENTS OF THE INVENTION
[0019] FIG. 1 shows how VF changes its morphology over time from
its onset. An ECG exhibiting VF is shown in its early stages
shortly after onset (left) and after 12 minutes (right). It can be
seen that at the beginning of VF higher frequencies, higher
amplitudes, higher slopes and a greater concentration of peaks are
found. These changes in VF morphology reflect the worsening
condition of the myocardium over time during uninterrupted VF.
[0020] The embodiments of the invention are based on the use of
so-called "quality markers" for VF. In the present context a VF
quality marker is a parameter, derived from an ECG exhibiting VF,
which represents the morphology of the ECG and, therefore, it
changes with the duration of the VF.
[0021] The embodiments teach a system incorporated into an
automated external defibrillator (AED) which measures the ECG of a
patient using the two defibrillator electrodes, calculates one or
more VF quality markers and estimates, using an algorithm, the
duration of the VF using the quality marker(s).
[0022] The prior art teaches two VF quality markers, referred to
herein as Frequency Ratio (FR) and Median Slope (MS). In addition,
two new VF quality markers are disclosed in the present
specification, referred to herein as Density and Amplitude of Peaks
(DA) and Average of Slopes (AS).
[0023] All, any or some of these quality markers can be used, in
conjunction with the defibrillator's own diagnostic algorithm, to
produce audible and visible indications to the operator to perform
CPR prior the delivery of the shock in order to increase the
chances of a successful resuscitation.
The VF Quality Markers
[0024] A sequence of ECG samples x.sub.0, x.sub.1, . . . ,
x.sub.N-1 in a window of N samples at fs samples per second is
processed in successive epochs in order to obtain the following
quality markers: Median Slope (MS), Average Slope (AS), Frequency
Ratio (FR) and Density and Amplitude of Peaks (DA).
The Median Slope
[0025] The median slope as a shock outcome predictor is disclosed
by Eilevstjonn et. al. (Eilevstjonn J, Kramer-Johansen J, Sunde K:
"Shock outcome is related to prior rhythm and duration of
ventricular fibrillation"; Resuscitation. 2007; 75 60-67).
[0026] The median slope (MS) is given by:
MS = median i = 1 , 2 , , N - 1 { ( x i - x i - 1 ) fs }
##EQU00001##
[0027] For simplicity, the slope here is denoted by
x.sub.i-x.sub.i-1 where two consecutive samples are used. However,
many more points can be used for calculating the slope. The slopes
are scaled and then sorted for the calculation of the median slope.
By definition, the median is the central value of an array of N
reordered samples. If N is odd, the value corresponding to the
position (N+1)/2 in the array is the median. Otherwise (if N is
even) the median is given by the semi-sum of the values in the
array corresponding to the positions N/2 and N/2+1.
The Average Slope
[0028] The average slope (AS) is expressed by:
AS = 1 N - 1 i = 1 N - 1 ( x i - x i - 1 ) fs ##EQU00002##
The Frequency Ratio
[0029] The Frequency Ratio (FR) as an indicator of VF duration is
disclosed by Sherman (Sherman L D: "The frequency ratio: An
improved method to estimate ventricular fibrillation duration based
on Fourier analysis of the waveform"; Resuscitation. 2006; 69:
479-486).
[0030] Sherman presented a method based on the frequency analysis
of the VF waveform. VF data was recorded for 12.5 minutes in 45
swine. The Fourier frequency spectra were calculated for 5 second
epochs. The average power at each frequency showed a marked loss of
frequencies above 8 Hz occurring at 5 min accompanied by an
increase in the power in frequency spectra from 3 to 5 Hz. The
Frequency Ratio was defined as the ratio of the power in the high
frequency band from 8 to 24 Hz compared to the power in the low
frequency band from 3 to 5 Hz. The Frequency Ratio was shown to
detect 90% of epochs in VF less than 5 min while allowing selection
of 74% of those epochs over 5 min. When the Frequency Ratio was set
to detect 90% of episodes of VF under 7 min, it was able to select
88% of those traces with VF over 7 min. The receiver operating
curve (ROC) for the frequency ratio had an area under the curve of
0.91 at 5 min and 0.95 at 7 min of VF duration.
[0031] Sherman claims that the Frequency Ratio is a strong
estimator of VF duration. However it is based on frequency analysis
which is computationally costly especially for the proposed range
of frequencies.
[0032] According to the procedure presented by Sherman, the
sequence of samples x.sub.0, . . . , x.sub.N-1 (in the time domain)
is transformed into the sequence of N complex numbers X.sub.0, . .
. , X.sub.N-1 (in the frequency domain) by the Discrete Fourier
Transform (DFT) according to the formula:
X k = n = 0 N - 1 x n - 2 .pi. N kn ##EQU00003##
k=0, 1, . . . , N-1 where
- 2 .pi. N kn ##EQU00004##
is a primitive N.sup.th root of unity.
[0033] Let P.sub.k=X.sub.k X.sub.k the power at the frequency index
k.
[0034] The low frequency band (3-5 Hz) is associated with the
frequency indexes from l.sub.1 to l.sub.n.
[0035] Then the power of the low frequency band is found by:
P low = k = l 1 l n P k ##EQU00005##
[0036] Similarly, the power of the high frequency band is given
by:
P high = k = h 1 h m P k ##EQU00006##
where h.sub.1 and h.sub.m are the frequency indexes for the high
frequency band (8-24 Hz).
[0037] Finally the Frequency Ratio (FR) is defined as:
F R = P high P low ##EQU00007##
[0038] However, due to the fact that Fast Fourier Transforms (FFTs)
are time consuming and use significant resources of the CPU, in the
present embodiment the estimation of the magnitudes for different
frequencies is carried out using integer filters--rather than
analyse for all frequencies, filter through those, at discrete
known frequencies, which are known to make the most significant
contribution. This technique is described in our Irish Patent
Application No. S2008/0785.
The Density and Amplitude of Peaks
[0039] For this quality marker an envelope of the ECG signal is
used. At every sample it is checked if a peak is detected which is
defined as an outstanding value outside the envelope. The envelope
is an artificial and auto adjusted signal created from the ECG
signal in order to contain it. However, an arriving sample, an
outstanding one, can lie outside the envelope. The principle in
deriving the envelope is to construct a ray aiming at the baseline
at a particular rate but the ray can be reset by a sample
"obstructing" its path to the baseline.
[0040] FIG. 2 is an example of an envelope for an ECG signal during
VF. The envelope 10 "contains" the signal 12 providing an
estimation of the maximum value of the signal at a particular
instant. This (local) maximum value or peak aims at the baseline
and is reset as the amplitude of the signal increases. As
mentioned, certain peaks 14 may lie outside (i.e. above) the
envelope.
[0041] In FIG. 3 the flow diagram for the calculation of the
envelope corresponding to each sample is presented. An explanation
of the variables used and their initial values in digital units is
presented in the following table:
TABLE-US-00001 Variable Description Initial value sample: ECG
sample that feeds the algorithm 0 at a rate of 170.6 samples/s
abs_sample: Absolute value of sample. 0 envelope: Artificially
created signal that acts as 5000 uV a ray aiming to the baseline.
In its path to the baseline it can be interrupted by an outstanding
sample that reset its height. max_peak: Propective value that
potentially 5000 uV "climb" to an expected maximum. last_peak:
Keeps a record of the value outside 5000 uV the envelope
encountered. seeker_decrement: Value taken from the peak_seeker to
8 uV give the following value for the peak_seeker. clearance:
Counter to control when max_peak 0 can be updated. rate Sample rate
(samples per second) 170.6
[0042] The parameters used in FIG. 2 and in the table above were
used with a sample rate of 170.6 samples per second. However the
parameters may change according to changes in the sample rate.
[0043] The Density of Peaks quality marker is given by:
DA = 1 N i = 0 N - 1 w i ##EQU00008##
where: w.sub.i=x.sub.i if x.sub.i is a peak lying outside
[0044] (i.e. above) the envelope, otherwise w.sub.i=0.
[0045] If desired, the formula for DA can be multiplied by fs (the
sample rate), as was done for the AS and MS quality markers. As fs
is a constant, the effect of this multiplication will be just
scaling but the original concept remains unchanged.
VF Duration
[0046] In order to compute an estimate of the VF duration from
these markers, the following model is used:
t = - B + C [ - ln ( Q - D A ) ] E ##EQU00009##
where t is the estimated VF duration in seconds, A, B, C, D and E
are parameters and Q is the value of the quality marker.
[0047] Q may represent more than one quality marker, and in general
a linear combination of the marker values may be used:
Q=p(AS)+q(MS)+r(DA)+s(FR)
where p, q, r and s are coefficients (scaling factors).
[0048] A, B, C, D, E, p, q, r and s are empirically derived from a
database of ECG signals during VF known to represent a range of ECG
qualities. For a particular ECG signal, a quality marker is
calculated at every sample. Using the quality marker, an estimation
of t is attempted and a comparison to the real time for VF duration
is carried out. The values for the parameters and coefficients are
iteratively adjusted with the aim of minimising the difference
between the estimated VF duration and the real one.
[0049] The following are the likely ranges of the constants A, B,
C, D and E: [0050] A=+1 to +1000 [0051] B=-500 to +500 [0052] C=+1
to +500 [0053] D=0 to +200 [0054] E=+1/2 to +5/2
Defibrillator Hardware
[0055] FIG. 4 is a block diagram of an automated external
defibrillator embodying the invention.
[0056] The hardware used in the embodiment is standard to an
automated external defibrillator and involves the measurement of
the ECG potentials through the two electrodes D1 which also serve
to deliver the shock therapy when required.
[0057] In use, the patient's ECG is sensed by the defibrillator
electrodes D1 and passed to a differential amplifier D2. The latter
is protected by circuitry D11 from the high voltage which, if
required by the patient, is applied to the defibrillation
electrodes D1 during electro-therapy. The resultant signal (.+-.3
mV) is passed to a first-order high pass filter D3 which, by means
of feedback into the differential amplifier D2, restores the DC
level to zero and removes the effects of respiration and movement
(below 1.6 Hz). The resultant signal is passed to a fourth-order
low pass filter D4 to remove mains pick-up and any other high
frequency noise (above 20 Hz). Finally, the signal is scaled in an
amplifier D5 to a level required by a microprocessor D6 for
analogue-to-WO digital conversion and sampling. In this embodiment
such sampling occurs at a rate of 170.6 samples per second.
[0058] The microprocessor D6 not only analyses the ECG signal,
using known techniques, to determine whether or not the patient is
in ventricular fibrillation, but controls indicators D7 which guide
the user in the delivery of the electro-therapy. The indicators may
comprise voice prompts and/or coloured lamps which are illuminated
to indicate predetermined conditions.
[0059] On detecting VF, the microprocessor D6 automatically
initiates a VF duration algorithm (to be described below) to
estimate the current phase of the VF. Depending on the outcome, the
defibrillator will either advise the application of CPR (to
increase the quality and therefore likely effectiveness of any
shock) or advise immediate shock. When a shock is advised, whether
immediate or after CPR, charging of capacitors D9 will be initiated
by the microprocessor D6 through activation of the charge circuit
D8. The voltage on the capacitors is sensed by the microprocessor
and, when at the correct energy level to be applied to the patient,
a bridge D10 is activated to apply a biphasic shock to the patient
when a "Shock" button (not shown) s pressed by the user.
[0060] It should be noted that the hardware shown in FIG. 4 is
largely standard to all automated external defibrillators. The
difference from the conventional machines is the ability to
estimate the duration of the VF and to advise the user via voice
prompts and/or coloured lamps D7 as to whether immediate
electro-therapy or CPR followed by electro-therapy is advised. This
is achieved using a novel algorithm embedded in the microprocessor
software.
Defibrillator Software
[0061] The steps of the software algorithm are:
[0062] If VF is detected: [0063] a. Determine the value Q of the
quality markers, or such of them as are used, for the preceding
4-second epoch of the ECG. Thereafter, for each digital sample,
update the value Q for the immediately preceding 4-second epoch of
the ECG. After the initial calculation of Q the calculation is an
updating process rather than a full calculation, in order to keep
processing time to a minimum. This ensures both avoid CPU overload
and the minimum delay in advising or administering the correct
therapy. In the present embodiment only the markers AS and DA are
calculated, and the value Q is given by:
[0063] Q=0.75(AS)+0.25(DA) [0064] b. At certain predefined
intervals, e.g. every 1 sec, for the current value of Q estimate
the duration t of VF from its onset using the foregoing model. In
this embodiment the following values were used in the model: [0065]
A=480 [0066] B=0 [0067] C=180 [0068] D=85 [0069] E=+1/2 [0070] c.
Determine the phase of VF. VF is deemed to be in the Electrical
Phase if t is lower than 240 seconds (4 min). [0071] d. Recommend
action to the user by audible and visual guidance and advice via
the indicator(s) D7: [0072] If VF is in the Electrical Phase,
immediate shock therapy is recommended and the defibrillator shock
circuits are enabled. [0073] If VF is not in the Electrical Phase,
CPR is recommended prior to shock therapy. The defibrillator shock
circuits are not enabled until it is recommended that CPR be given,
but may be manually overridden.
[0074] FIG. 5 is an alternative algorithm for determining the VF
phase in the case of the DA quality marker. As before, for each
4-second epoch the envelope 10 is calculated as described with
reference to FIGS. 2 and 3, step 100. In steps 110 and 112 the DA
quality marker is determined according to the formula for DA given
earlier as the average of the absolute values of the samples in the
epoch that lay above the envelope.
[0075] Next, however, a direct estimation of the VF phase is made
by comparing the value of DA with a predetermined threshold level,
step 114. In the present embodiment this threshold is 0.29
mV/sample, and if DA is equal or greater than this threshold VF is
assumed to be in the Electrical Phase. Thus immediate shock therapy
is recommended and the defibrillator shock circuits are
enabled.
[0076] If DA is less than the threshold, it is assumed that VF is
not in the Electrical Phase. Thus CPR is recommended prior to shock
therapy. The defibrillator shock circuits are not enabled until it
is recommended that CPR be given, but may be manually
overridden.
[0077] It will be seen that this embodiment avoids the need to
estimate the duration of VF.
[0078] The invention is not limited to the embodiments described
herein which may be modified or varied without departing from the
scope of the invention.
* * * * *